Across the rolling steppes of Inner Mongolia, a dramatic transformation is underway. Farmers harvesting potatoes now share their landscape with massive rectangular structures—data centres that will power China's artificial intelligence platforms. Ulanqab, historically known as the nation's potato capital, has become an unlikely digital frontier, with hundreds of construction workers in yellow helmets assembling the infrastructure that sits at the heart of Beijing's technological competition with the United States.
The city has signed 89 data centre projects involving 68 enterprises by the end of June 2026, with total investment exceeding 500 billion yuan. According to industry reports, the build-out has been described as reaching more than 5 million server racks, with the planned capacity estimated at about 12.5 GW of computing power already in operation and planned combined.
On construction sites across the region, welding sparks fly from metal scaffolding as cranes move with precision to build rooftops and workers balance on ladders to finish brickwork.
"This is China speed,"jokes one worker, describing his country as
"an infrastructure maniac."An electrical engineer explains the scale of ambition:
"There is one phase every year. This is the second phase. There will be many more phases afterwards."
Why is Ulanqab becoming China's AI hub?
The city's emergence as a computing centre reflects strategic decisions by Beijing. The region offers abundant renewable energy capacity—wind and solar installations already built across the province—while land costs remain low and cooler weather reduces water demand for cooling systems. To the Chinese Communist Party, these data centres are critical infrastructure for economic and national security, essential to embedding artificial intelligence across 90% of industries and society by 2030.
Major technology firms including Huawei, ByteDance and DeepSeek have established operations here. Envision's Galaxy Campus in Ulanqab began operations in August 2026 as a 2 GW AI campus powered by renewables, with a first phase already commissioned. ByteDance was reportedly seeking 5–6 GW of additional compute in Ulanqab, targeting delivery by early 2028, while DeepSeek was reported to be planning about 1 GW of compute in the city.
Data centre and related internet services in Ulanqab consumed 4 billion kWh in the first half of 2026, up 85.2% year on year, underscoring the rapid acceleration of the build-out.
How is China attracting talent to compete with the US?
Beyond infrastructure, Beijing is cultivating a pool of young engineers and researchers who increasingly view their country as a place of opportunity. Tensions with the United States and restrictions on immigration have altered the calculus for many scientists and technologists.
"Fewer students and researchers are heading to the US, and more are returning from the US to China. That is very clear,"says Jack Zhang from the University of Science and Technology of China, who himself was recruited by a high-level programme after studying at Yale and Cambridge.
Zhang explains his own motivation:
"I wanted to take some of the AI work I was doing in the lab and apply it to human health and medicine. That requires more resources, more people and sometimes more government support, which can be difficult to access overseas."
China produces around five million science and technology graduates each year, compared to around half a million in the United States. As more of them remain in China, they encounter a growing ecosystem for innovation. AI developer Dang Ronghao, 27, who works for a leading technology company, observes that Chinese campuses are becoming more open, allowing him to collaborate with researchers worldwide from his base in Shanghai.
"I've seen research combining AI with nuclear fusion, AI with astrophysics, AI with new materials, and AI with the development of new drugs. I think we are entering a good era."
What is China's alternative approach to AI development?
Rather than simply attempting to catch up with American firms, Beijing is promoting a distinct model. Xi Jinping himself is encouraging Chinese AI companies to develop more affordable, open-source models that allow people, companies and countries to modify them as they wish. This contrasts sharply with the closed-model approaches used by developers including OpenAI and Anthropic, whose design elements remain proprietary secrets.
Beijing favours this strategy partly because it has proven effective. By sharing technology, Chinese AI firms have been able to innovate rapidly and overcome one of their biggest obstacles: limited access to advanced chips that would unlock greater computing power. The approach also reflects Beijing's broader ambition to shape global AI governance. Xi has repeatedly called for more regulation and a global framework for others to follow.
The state-run Global Times, which reflects official thinking, has dismissed American concerns about Chinese AI advancement as part of a
"Cold War playbook"designed to
"contain"China's rise. Meanwhile, President Trump has stated that
"whoever wins AI, wins,"while the CEO of AI firm Anthropic, Dario Amodei, has called for maintaining the US-led ban on selling advanced chips and equipment to China, warning that a Chinese lead would
"pose grave danger for the US and the world."
What are the environmental and social costs?
The rapid expansion of data centre infrastructure has created tensions in rural communities. China aims for most of its data centres to be fuelled by green energy by 2030, yet some of the grid remains powered by coal. Water supply remains a concern in the arid regions of Inner Mongolia, where data centres require significant cooling resources.
Entire villages have been relocated to make way for construction, though the precise number of affected residents and compensation offered remain difficult to verify. Few locals will discuss the topic openly, as voicing concerns can be risky.
"This big data has absolutely nothing to do with ordinary people,"says one resident, speaking anonymously.
"What benefits has it brought us?"He reports that few locals have found employment at the facilities, while others express fears that their land will be seized and they will lose their ability to farm for a living.
Not all voices are critical. A man identified as Zhang, standing alongside potato pickers in a field near the data centres, expressed optimism about China's technological trajectory.
"China is now at the forefront of the world,"he says, noting that his children have managed to secure jobs at the new facilities.
"It's not like the 1960s, when it was relatively backward. China is doing alright now. Whether in economic development or other areas, it is doing well."
Surveys suggest there is relatively little public opposition to AI or its widespread adoption compared to many other countries, though evaluating such figures in a heavily-censored environment remains challenging. Beijing has been explicit that it views AI as a solution to its ageing workforce, rapidly automating manufacturing and production. This is generating anxiety over potential job losses, which again is difficult to measure accurately.
Some observers argue that China still prioritises conformity over intellectual challenge.
"It's not as open as I wish,"says Professor Liu Jun from the Department of Statistics and Data Science at Tsinghua University, who has studied at both Harvard and Stanford.
"I keep encouraging students here to ask more questions, to challenge more, but that's hard to change. I think partly it's culture, partly it's the system."
How does this affect ordinary people in rural areas?
Far from the laboratories where researchers grapple with AI's frontier, many Chinese citizens face more immediate economic pressures: a sluggish economy, rising youth unemployment, weak consumer spending and property prices that have plummeted. In rural areas, an ageing population compounds these challenges.
One farmer in Inner Mongolia, tending his sheep from a cart, expresses scepticism about the technology boom.
"Big data and all that… pretty good – but it has nothing to do with me,"he says. He has little use for artificial intelligence and barely uses the internet. His phone screen is cracked, and he pays just over 40 yuan monthly for wifi from a pocket dongle.
"There haven't really been any changes for me. I don't know about the future,"he reflects.
This kind of scepticism presents a challenge for Beijing. Enormous change is occurring across the country, yet it remains unclear how many ordinary citizens share their leaders' enthusiasm for this technology-driven future.
What happens next in US-China AI competition?
The stakes of this technological race are being tested at the highest diplomatic levels. Trump and Xi are scheduled to meet on 24 September 2026, with artificial intelligence safety expected to feature prominently on the agenda. US and Chinese officials discussed a proposed AI safety notification mechanism during talks in New York on 20 September, ahead of the summit.
A proposed US-China AI dialogue and notification mechanism may be considered at that summit if both sides accept the idea. With calls for AI safety growing louder, this week's summit at the White House carries special significance. Whether the two sides can collaborate on regulating AI—and whether any meaningful slowdown in development is realistic—remain open questions. For now, in places like Ulanqab, China's AI ambitions continue to take physical shape, transforming rural landscapes into the computing infrastructure that will define technological competition for years to come.
